Clouds and the Earth's Radiant

Energy System

Quick-Look Results - Data Validation

View GIF         View PDF

View EPS Comparison of CERES 8-12 micron window unfiltered radiance with GOES 10.7micron window radiance
GOES-8 and GOES-9 10.7 micron window radiances are used to validate CERES 8-12 micron window unfiltered radiances for data collected during December 29, 1997. All three sets of data are spatially gridded into the CERES 1-degree equal-angle grid system. CERES window data are also temporally gridded into 24 hourly-boxes to allow for better time-matching with the GOES data set. Data with viewing zenith angle larger than 60 degree are not processed to limit limb-darkening effects. Correlation analyses are performed to assess the relationship between CERES window and the GOES-8/GOES-9 window data. The results (shown in the scatter diagram) for both CERES vs. GOES-8 and CERES vs. GOES-9 comparison are very good with correlation coefficient of 0.929 and 0.937, respectively. These values are consistent with comparisons (Minnis et al., 1991) made earlier between ERBE scanner Broadband radiances and GOES window radiances. The scatter in the data are caused by mismatches in both time and space between the CERES and the GOES data sets. Additional analyses will be performed for GMS and METEOSAT window data as they become available in the future. For other information, please contact Takmeng Wong at takmeng.wong@larc.nasa.gov.

View GIF         View PDF

View EPS CERES Deep Space Calibration Maneuver
The Deep Space Calibration Maneuver was performed on January 7 and 8, 1998. During deep space calibration, the instrument was pitched to view the cold space over the entire scan of Earth viewing positions. The measurements are taken with the instrument operating in both cross-track and rotating azimuth plane (RAPS) modes. This plot shows the sensor output from total channel in cross-track scan mode during the first orbit of deep space maneuver.
The axes are sample number (elevation scans across the Earth), scan number (6.6 second scan time), and the total channel measurement in counts. The ragged peaks at the beginning and end of the period are partial scans of the Earth while the vertical bar in the middle of the scan shows the internal calibration source at ambient temperatures. The near-zero values elsewhere show the sensor response while viewing deep space. These measurements are used to help calibrate CERES.

View GIF         View PDF

View EPS Validation of Convolution Algorithms
This figure shows the area covered by the first 4500 CERES footprints, or fields-of-view (FOV), in hour 00, January 3, 1998, for which VIRS imager data were available to support convolution and inclusion in the hour 00 Single Satellite Footprint (SSF).
During this interval, the scan mode of the CERES instrument was switched from rotating azimuth plane scan (RAPS) to cross-track, and the transition between modes is evident from the pattern of the plotted footprints. Due to the overlap between successive CERES samples, individual footprints cannot be discriminated in the figure. Therefore, each "stripe" in the plot represents the envelope of a given scan, consisting of a sequence of overlapping footprints.
The specific data plotted is effective pressure of the lowest cloud layer determined for each VIRS pixel by the cloud retrieval algorithms in CERES Subsystem 4.3, and convolved with the CERES FOV by the "cookie-cutter" algorithm in Subsystem 4.4. The cloud effective pressure is inversely related to cloud height. The plot may thus be viewed as a type of false-color cloud picture, similar to conventional high-resolution cloud imagery from weather satellites.
In the plot, the color values are assigned in 100-hPa intervals to pixels for which the cloud property retrieval algorithms determined that cloud was present and calculated an effective pressure value. Cloud-free pixels and those for which the cloud algorithms could not make a reliable determination are plotted in white.
Depending on the viewing zenith angle and azimuth, the number of VIRS pixels convolved into each CERES FOV varies from about 70 near nadir to over 4000 in high-zenith along-track RAPS footprints.
Due to the low altitude of TRMM and the relatively small angular cross-section of the CERES point spread distribution, successive cross-track scans do not overlap near nadir. This can be seen in the slight voids between scanlines, visible in the sequence of about 8.5 cross-track scan cycles in the right side of the plot.
This plot program is useful in validation of the imager-derived cloud property data as well as in verifying the functionality of the convolution algorithms, including the CERES point spread function.

View GIF         View PDF

View EPS TRMM ANALYSIS
One component of CERES is the use of satellite measurements in combination with meteorological analyses and radiative transfer modeling to estimate the full vertical profile of radiative fluxes from the surface to the top of the atmosphere (TOA). The vertical profile of fluxes is called the Surface and Atmospheric Radiation Budget (SARB). Here we show how the process to determine the SARB adjusts the column-integrated water vapor (precipitable water or PW).
CERES measurements of longwave (LW) radiance and an assumed Angular Distribution Model (ADM) yield an observation of outgoing LW radiation (OLR). The OLR is also computed in CERES using NASA Goddard's GEOS-1.3 global meteorological analysis for temperature and humidity and the Fu-Liou radiative transfer code. In the upper left panel, the OLR for a clear sky swath, the difference of observed OLR minus computed OLR (UNTUNED TRMM-MODEL LW) is positive and quite large; up to 30 W/m**2 near 10N, 265E (due south of Chiapas, Mex.).
In this case, the CERES CRS algorithm then constrains the computed OLR to more closely agree with the observed OLR. This is done by decreasing the atmospheric PW and increasing the sea surface (skin) temperature (SST). Here we focus on the the adjustments to PW, for which there is independent data. The lower left panel shows the PW field (MOA/GEOS-1.3 PW) used by CERES in the initial ("untuned") calculation of OLR; note the region near 10N, 265E, where the PW used by CERES is ~ 5 cm. In the upper right panel, the CRS constrainment algorithm reduces the PW (ADJUSTMENT OF PW) in this region by ~ 1.5 cm. The final PW near 10N, 265E would be 5 cm - 1.5 cm = 3.5 cm, but is the adjusted PW then correct? We think it's pretty good. In the lower right panel, a completely independent PW (SSMI PW) also shows values of ~ 3.5 cm near 10N, 273W. The independent PW (lower right) was obtained from the DMSP SSM/I microwave instrument and was not used by Goddard in producing the GEOS-1.3 PW (lower left).
After adjusting the PW, the re-computed OLR (not shown) is still quite different than the OLR observed by the CERES instrument. For example, near 10N, 265E the differences of re-computed and observed OLR are ~20 W/m**2.

View GIF         View PDF

View EPS Comparison of CERES Measurements with Theory
All three CERES channels (Total, Window, and Shortwave) register signals of thermal infrared radiation emitted by the earth and the atmosphere. Even the Shortwave channel, which is designed to measure reflected sunlight, detects a small amount of emitted thermal radiation. The plot compares measured Shortwave and Window signals from a large number of CERES footprints at night (when there is no reflected solar radiation). A theoretical curve is dashed; this is based on the Air Force MODTRAN3 band model and climatological atmospheres. Measurement and theory are in rough agreement. The MODTRAN3 model uses similar, but more detailed, radiative transfer physics to that used in climate models. CERES data will be used to test climate models.
CERES aims to clearly delineate the separate contributions of reflected solar radiation and emitted terrestrial thermal radiation to the total Earth Radiation Budget. This requires a careful analysis of the optical physics of the CERES measurement channels. The relation shown in the figure will be tested and refined for the analysis.

View GIF         View PDF

View EPS CERES image:
This plot contains 20 scanlines of total radiance from western Mexico processed through the Coastline Detection geolocation assessment algorithm. Coastline detections are plotted as circular symbols. This sample contains a 0.9 km cross-track bias and a 2.2 km along-track error. Many samples are collected to determine systematic biases.

View GIF         View PDF

View EPS VIRS image:
This plot contains 441 scanlines of channel 5 brightness temperature from western Mexico processed through the Coastline Detection geolocation assessment algorithm. Coastline detections are plotted as circular symbols. This sample contains a 0.38 km cross-track bias and a 0.13 km along-track error. Many samples will be collected to determine systematic biases.

Return
to
Quick-Look
Return
to
CERES

Responsible NASA Officials: Dr. Bruce R. Barkstrom, Principal Investigator of CERES Instrument and Dr. Bruce A. Wielicki, Principal Investigator of CERES Interdisciplinary Science Team

Page Last Updated April 10, 1998, 11:00

Web Curator: Kay Costulis

View GIF View PDF View EPS		Comparison of CERES 8-12 micron window unfiltered radiance with GOES 10.7micron window radiance GOES-8 and GOES-9 10.7 micron window radiances are used to validate CERES 8-12 micron window unfiltered radiances for data collected during December 29, 1997. All three sets of data are spatially gridded into the CERES 1-degree equal-angle grid system. CERES window data are also temporally gridded into 24 hourly-boxes to allow for better time-matching with the GOES data set. Data with viewing zenith angle larger than 60 degree are not processed to limit limb-darkening effects. Correlation analyses are performed to assess the relationship between CERES window and the GOES-8/GOES-9 window data. The results (shown in the scatter diagram) for both CERES vs. GOES-8 and CERES vs. GOES-9 comparison are very good with correlation coefficient of 0.929 and 0.937, respectively. These values are consistent with comparisons (Minnis et al., 1991) made earlier between ERBE scanner Broadband radiances and GOES window radiances. The scatter in the data are caused by mismatches in both time and space between the CERES and the GOES data sets. Additional analyses will be performed for GMS and METEOSAT window data as they become available in the future. For other information, please contact Takmeng Wong at takmeng.wong@larc.nasa.gov.

View GIF View PDF View EPS		CERES Deep Space Calibration Maneuver The Deep Space Calibration Maneuver was performed on January 7 and 8, 1998. During deep space calibration, the instrument was pitched to view the cold space over the entire scan of Earth viewing positions. The measurements are taken with the instrument operating in both cross-track and rotating azimuth plane (RAPS) modes. This plot shows the sensor output from total channel in cross-track scan mode during the first orbit of deep space maneuver. The axes are sample number (elevation scans across the Earth), scan number (6.6 second scan time), and the total channel measurement in counts. The ragged peaks at the beginning and end of the period are partial scans of the Earth while the vertical bar in the middle of the scan shows the internal calibration source at ambient temperatures. The near-zero values elsewhere show the sensor response while viewing deep space. These measurements are used to help calibrate CERES.

View GIF View PDF View EPS		Validation of Convolution Algorithms This figure shows the area covered by the first 4500 CERES footprints, or fields-of-view (FOV), in hour 00, January 3, 1998, for which VIRS imager data were available to support convolution and inclusion in the hour 00 Single Satellite Footprint (SSF). During this interval, the scan mode of the CERES instrument was switched from rotating azimuth plane scan (RAPS) to cross-track, and the transition between modes is evident from the pattern of the plotted footprints. Due to the overlap between successive CERES samples, individual footprints cannot be discriminated in the figure. Therefore, each "stripe" in the plot represents the envelope of a given scan, consisting of a sequence of overlapping footprints. The specific data plotted is effective pressure of the lowest cloud layer determined for each VIRS pixel by the cloud retrieval algorithms in CERES Subsystem 4.3, and convolved with the CERES FOV by the "cookie-cutter" algorithm in Subsystem 4.4. The cloud effective pressure is inversely related to cloud height. The plot may thus be viewed as a type of false-color cloud picture, similar to conventional high-resolution cloud imagery from weather satellites. In the plot, the color values are assigned in 100-hPa intervals to pixels for which the cloud property retrieval algorithms determined that cloud was present and calculated an effective pressure value. Cloud-free pixels and those for which the cloud algorithms could not make a reliable determination are plotted in white. Depending on the viewing zenith angle and azimuth, the number of VIRS pixels convolved into each CERES FOV varies from about 70 near nadir to over 4000 in high-zenith along-track RAPS footprints. Due to the low altitude of TRMM and the relatively small angular cross-section of the CERES point spread distribution, successive cross-track scans do not overlap near nadir. This can be seen in the slight voids between scanlines, visible in the sequence of about 8.5 cross-track scan cycles in the right side of the plot. This plot program is useful in validation of the imager-derived cloud property data as well as in verifying the functionality of the convolution algorithms, including the CERES point spread function.

View GIF View PDF View EPS		TRMM ANALYSIS One component of CERES is the use of satellite measurements in combination with meteorological analyses and radiative transfer modeling to estimate the full vertical profile of radiative fluxes from the surface to the top of the atmosphere (TOA). The vertical profile of fluxes is called the Surface and Atmospheric Radiation Budget (SARB). Here we show how the process to determine the SARB adjusts the column-integrated water vapor (precipitable water or PW). CERES measurements of longwave (LW) radiance and an assumed Angular Distribution Model (ADM) yield an observation of outgoing LW radiation (OLR). The OLR is also computed in CERES using NASA Goddard's GEOS-1.3 global meteorological analysis for temperature and humidity and the Fu-Liou radiative transfer code. In the upper left panel, the OLR for a clear sky swath, the difference of observed OLR minus computed OLR (UNTUNED TRMM-MODEL LW) is positive and quite large; up to 30 W/m2 near 10N, 265E (due south of Chiapas, Mex.). In this case, the CERES CRS algorithm then constrains the computed OLR to more closely agree with the observed OLR. This is done by decreasing the atmospheric PW and increasing the sea surface (skin) temperature (SST). Here we focus on the the adjustments to PW, for which there is independent data. The lower left panel shows the PW field (MOA/GEOS-1.3 PW) used by CERES in the initial ("untuned") calculation of OLR; note the region near 10N, 265E, where the PW used by CERES is ~ 5 cm. In the upper right panel, the CRS constrainment algorithm reduces the PW (ADJUSTMENT OF PW) in this region by ~ 1.5 cm. The final PW near 10N, 265E would be 5 cm - 1.5 cm = 3.5 cm, but is the adjusted PW then correct? We think it's pretty good. In the lower right panel, a completely independent PW (SSMI PW) also shows values of ~ 3.5 cm near 10N, 273W. The independent PW (lower right) was obtained from the DMSP SSM/I microwave instrument and was not used by Goddard in producing the GEOS-1.3 PW (lower left). After adjusting the PW, the re-computed OLR (not shown) is still quite different than the OLR observed by the CERES instrument. For example, near 10N, 265E the differences of re-computed and observed OLR are ~20 W/m2.

View GIF View PDF View EPS		Comparison of CERES Measurements with Theory All three CERES channels (Total, Window, and Shortwave) register signals of thermal infrared radiation emitted by the earth and the atmosphere. Even the Shortwave channel, which is designed to measure reflected sunlight, detects a small amount of emitted thermal radiation. The plot compares measured Shortwave and Window signals from a large number of CERES footprints at night (when there is no reflected solar radiation). A theoretical curve is dashed; this is based on the Air Force MODTRAN3 band model and climatological atmospheres. Measurement and theory are in rough agreement. The MODTRAN3 model uses similar, but more detailed, radiative transfer physics to that used in climate models. CERES data will be used to test climate models. CERES aims to clearly delineate the separate contributions of reflected solar radiation and emitted terrestrial thermal radiation to the total Earth Radiation Budget. This requires a careful analysis of the optical physics of the CERES measurement channels. The relation shown in the figure will be tested and refined for the analysis.

View GIF View PDF View EPS		CERES image: This plot contains 20 scanlines of total radiance from western Mexico processed through the Coastline Detection geolocation assessment algorithm. Coastline detections are plotted as circular symbols. This sample contains a 0.9 km cross-track bias and a 2.2 km along-track error. Many samples are collected to determine systematic biases.

View GIF View PDF View EPS		VIRS image: This plot contains 441 scanlines of channel 5 brightness temperature from western Mexico processed through the Coastline Detection geolocation assessment algorithm. Coastline detections are plotted as circular symbols. This sample contains a 0.38 km cross-track bias and a 0.13 km along-track error. Many samples will be collected to determine systematic biases.